VISTA Variables in the Vía Láctea (VVV): Halfway Status and Results

Astronomical Science

Maren Hempel1,2 I. Félix Mirabel36,37 12 Departamento de Física, Universidad Dante Minniti1,3 Lorenzo Monaco8 de La Serena, Chile István Dékány1 Christian Moni-Bidin38 13 Department of Astronomy, University of Roberto K. Saito1,4 Lorenzo Morelli39 Florida, Gainesville, USA Philip W. Lucas5 Nelson Padilla1 14 Universidade de São Paulo, Brazil Jim P. Emerson6 Tali Palma7 15 Facultad de Ciencias Astronómicas y Andrea V. Ahumada7,8,9 Maria Celeste Parisi7 Geofísicas, Universidad Nacional de Suzanne Aigrain10 Quentin Parker40,41 La Plata, Argentina Maria Victoria Alonso7 Daniela Pavani17 16 Space Telescope Science Institute, Javier Alonso-García1 Pawel Pietrukowicz42 Baltimore, USA Eduardo B. Amôres11 Grzegorz Pietrzynski24,43 17 Universidade Federal do Rio Grande do Rodolfo Angeloni1 Giuliano Pignata44 Sul, Porto Alegre, Brazil Julia Arias12 Marina Rejkuba8 18 Departamento de Física y Astronomía, Reba Bandyopadhyay13 Alejandra Rojas1 Facultad de Ciencias, Universidad de Rodolfo H. Barbá12 Alexandre Roman-Lopes12 Valparaíso, Chile Beatriz Barbuy14 Maria Teresa Ruiz19 19 Departamento de Astronomía, Univer- Gustavo Baume15 Stuart E. Sale1,18,45 sidad de Chile, Santiago, Chile Juan Carlos Beamin1 Ivo Saviane8 20 Institute for Astronomy, The University Luigi Bedin16 Matthias R. Schreiber18 of Edinburgh, UK Eduardo Bica17 Anja C. Schröder46,47 21 Joint Astronomy Centre, Hilo, USA Jordanka Borissova18 Saurabh Sharma18 22 Kavli Institute for Astronomy and Leonardo Bronfman19 Michael Smith48 Astrophysics, Peking University, China Giovanni Carraro8 Laerte Sodré Jr.14 23 Astrophysics Group, Imperial College Márcio Catelan1 Mario Soto12 London, UK Juan J. Clariá7 Andrew W. Stephens49 24 Departmento de Astronomía, Universi- Carlos Contreras1 Motohide Tamura50 dad de Concepción, Chile Nicholas Cross20 Claus Tappert18 25 Max Planck Institute for Astronomy, Christopher Davis21 Mark A. Thompson5 Heidelberg, Germany Richard de Grijs22 Ignacio Toledo51 26 Department of Astrophysics, University Janet E. Drew5,23 Elena Valenti8 of Oxford, UK Cecilia Fariña15 Leonardo Vanzi52 27 Department of Physics, University of Carlos Feinstein15 Walter Weidmann7 Cincinnati, USA Eduardo Fernández Lajús15 Manuela Zoccali1 28 School of Physics & Astronomy, Univer- Stuart Folkes5,18 sity of Leeds, UK Roberto C. Gamen15 29 Institute of Astronomy, University of Douglas Geisler24 1 Instituto de Astrofísica, Pontificia Cambridge, UK Wolfgang Gieren24 Universidad Católica de Chile, Chile 30 Jodrell Bank Centre for Astrophysics, Bertrand Goldman25 2 The Milky Way Millennium Nucleus, The University of Manchester, UK Oscar González8 Santiago, Chile 31 NASA-Ames Research Center, Moffett Andrew Gosling26 3 Vatican Observatory, Italy Field, USA Guillermo Gunthardt12 4 Universidade Federal de Sergipe, Depar- 32 Instituto de Astrofísica de Canarias, Sebastian Gurovich7 tamento de Física, Cristóvão, Brazil La Laguna, Tenerife, Spain Nigel C. Hambly20 5 Centre for Astrophysics Research, 33 Department of Physics, Texas Tech Margaret Hanson27 Science and Technology Research University, Lubbock, USA Melvin Hoare28 Institute, University of Hertfordshire, 34 Istituto di Astrofisica Spaziale e Fisica Mike J. Irwin29 Hatfield, UK Cosmica di Bologna, Italy Valentin D. Ivanov8 6 Astronomy Unit, School of Physics and 35 Rochester Institute of Technology, Andrés Jordán1 Astronomy, Queen Mary University of Rochester, USA Eamonn Kerins30 London, UK 36 Service d’Astrophysique – IRFU, Karen Kinemuchi31 7 Observatorio Astronómico de Córdoba, CEA-Saclay, Gif sur Yvette, France Radostin Kurtev18 Argentina 37 Instituto de Astronomía y Física del Andy Longmore20 8 ESO Espacio, Buenos Aires, Argentina Martin López-Corredoira32 9 Consejo Nacional de Investigaciones 38 Instituto de Astronomía, Universidad Tom Maccarone33 Científicas y Técnicas, Buenos Aires, Católica del Norte, Antofagasta, Chile Eduardo Martín32 Argentina 39 Dipartimento di Astronomia, Universitá Nicola Masetti34 10 School of Physics, University of Exeter, di Padova, Italy Ronald E. Mennickent24 UK 40 Department of Physics, Macquarie Uni- David Merlo7 11 SIM, Faculdade de Ciências da versity, Sydney, Australia Maria Messineo35 Universidade de Lisboa, Portugal

24 The Messenger 155 – March 2014 41 Australian Astronomical Observatory, 1.1 by 1.5 degree field of view, ideal for Epping, Australia surveys covering many hundreds of 42 Nicolaus Copernicus Astronomical square degrees. The large field of view in Center, Warsaw, Poland combination with the spatial resolution P93+ P85 360 hr (18.6%) 300 hr (15.6%) 43 Warsaw University Observatory, of 0.339 arcseconds per pixel make Warsaw, Poland VISTA/VIRCAM an ideal instrument to 44 Departamento de Ciencias Físicas, observe the most crowded and extincted Universidad Andres Bello, Santiago, regions of the Milky Way, i.e. the central P87 Chile regions of the Bulge and the mid-plane 292 hr (15.1%) 45 Rudolf Peierls Centre for Theoretical regions of the Galactic Disc. The feasible P91 Physics, Oxford, UK observing period for the VVV survey is 550 hr (28.5%) 46 SKA/KAT, Cape Town, South Africa limited to six months, between February P89 47 Hartebeesthoek Radio Astronomy and October, requiring careful scheduling 275 hr (14.3%) Observatory, Krugersdorp, South of the individual observing periods. P90 Africa 152 hr ( 7.9 %) 48 The University of Kent, Canterbury, UK The first main phase of the survey, 49 Division of Optical and Infrared Astron- consisting of the YZJHKs multi-colour omy, National Astronomical Observa- observations, was assigned to the first Figure 1. Observing schedule for the VVV survey in tory of Japan, Tokyo, Japan semester of the survey (Period 85, 2010) comparison with the total allotted observing time of 50 1929 hours. Note, that ca. 50% of the P91 observa- Gemini Observatory, Hawaii, USA to obtain a first overview of the survey tions were delayed due to poor weather. The remain- 51 ALMA Observatory, Santiago Chile area (described in Section 2). The vast ing 360 hours will be distributed over Period 93 52 Departamento de Ingeniería Eléctrica, majority of the observations however form (2014) and the following years. Pontificia Universidad Católica de Chile, the variability campaign, which started in Santiago, Chile parallel with the multi-colour observations Already, after the first two observing in Period 85, but then occupied all the fol- periods, despite the modest amount of lowing observing periods. For the variabil- data (see Figure 1) it has become obvious The VISTA Variables in the Vía Láctea ity campaign, VVV was allotted 300 hours that the data reduction, analysis and (VVV) survey is one of six near-infrared in 2010 (including multi-colour observa- even data storage/transfer are daunting ESO public surveys, and is now in its tions), 292 hours in 2011, 275 hours in tasks. These considerations affected fourth year of observing. Although far 2012 and 702 hours in 2013 (split between not only the survey team, conducting a from being complete, the VVV survey Periods 90 and 91). The remaining 360 wide range of science projects, final qual- has already delivered many results, hours of the survey will be used not only to ity control, preparation of observations some directly connected to the intended gather additional data for the variability and the public data release, but also the science goals (detection of variable survey (see below), but also to aid the Cambridge Astronomical Survey Unit stars, microlensing events, new star ongoing proper motion study on the Solar (CASU), in charge of the pipeline reduc- clusters), others concerning more exotic Neighbourhood and in searching for tion of all data, as well as ESO, perform- objects, e.g., novae. Now, at the end of microlensing events in a selected Bulge ing the observations, the first level of the fourth observing period, and com- area. The long-term status of the VVV sur- quality control and the final data release. prising roughly 50% of the proposed vey, especially in combination with other observations, the status of the survey, data (e.g., 2MASS and WISE) make it very as well some of results based on the suitable to conduct proper motion studies, Multi-colour photometry VVV data, are presented. which are the subject of a later section. The multi-colour observations in the five The following observations were sched- broadband filtersY , Z, J, H and Ks com- Introduction uled for each period (Minniti et al., 2010): menced in March 2010 and were con- – P85 (April–October 2010): YZJHKs cluded by September 2011. Starting with The Visible and Infrared Survey Tele- and additional Ks-band observations the J-, H- and Ks-band data only, an scope for Astronomy (VISTA, Emerson & for the whole survey area; unprecedented first multi-colour view on Sutherland, 2010) has been operated by – P87 (April–October 2011): Ks-band the inner region of the Milky Way Bulge ESO for four years. Observations began observations of the complete survey composed of 84 million stellar sources with Science Verification in September area; was presented by Saito et al. (2012a; see 2009, and this was followed by the six – P89 (April–October 2012): main varia also the ESO Release 12421), followed by public surveys, one of which is VISTA bility campaign on the Bulge area; the study of Soto et al. (2013) of 88 mil- Variables in the Via Láctea (see Minniti et – P90 (November 2012 – March 2013): lion stellar sources in the southern Galac- al., 2010; Saito et al., 2010; Catelan et al., variability campaign on the outer Disc tic Disc. 2011). The only science instrument cur- area; rently available at VISTA is the wide-field – P91: (April–October 2013): variability Multi-colour observations with high spa- camera VIRCAM (Dalton et al., 2006; campaign of the inner Disc region and tial resolution and photometric depth, like Emerson & Sutherland, 2010), offering a the Bulge. the VVV datasets, are not only valuable

The Messenger 155 – March 2014 25 Astronomical Science Hempel M. et al., VISTA Variables in the Vía Láctea (VVV)

, RSDKK@QRNTQBDR Figure 2. The source density in the Bulge region of the VVV survey is shown. Only point sources which were detected in all five bands are included in this plot. Empty fields represent data that did not fulfill the stringent quality criteria and are not part of the public data release. The latter observations will be repeated in due course. Red Clump (RC) stars are an ideal tracer of the reddening effect, since their colour l and luminosity are well defined and the l dependence on age and metallicity well DF understood. Gonzales et al. (2013) used l !C the distribution of the RC stars within l the VVV Bulge area to derive a reddening l map covering the Galactic Bulge (see Figure 3 of Gonzalez et al. [2013]) within l 4 degrees of the Galactic Plane with a l spatial resolution of 2 arcminutes. At l larger Galactic latitudes, where the reddening varies on a larger scale, the reso- l lution is slightly lower, but still ≤ 6 arc- l minutes. Such a high spatial resolution is, l l l l l for instance, essential when studying +CDF stellar populations in Milky Way star clusters, which, depending on their position, probes of the Galactic structure, but be less reliable for the most reddened are affected by differential reddening. also extremely important to tackle one of regions near the Galactic Plane, where the large problems in Galactic studies, an extinction in the V magnitude (AV) of Using the RC stars as bona fide dis- the effect of reddening. Line of sight 30 mag can be reached. Including the tance indicators of the innermost region reddening hampers all photometric single epoch observations in the Y- and of the Bulge along various lines of sight, methods, such as for age and metallicity Z-bands reveals the distribution of the Gonzalez et al. (2011) could also study determination. Detailed reddening maps, dust obscuring the inner regions of the inner Galactic Bar, and provide addi- such as the ones provided by Schlegel the Bulge, and enables the high stellar tional evidence for the existence of a sec- et al. (1998) are widely used to correct for density to become clearly visible (see ondary bar structure, as suggested by Galactic reddening, but were known to Figure 2). Nishiyama et al. (2005). In a similar way, Saito et al. (2011), based on 2MASS data, mapped the X-shaped structure of the Bulge, showing extension far beyond the spatial extent of the inner Bar structure.

Variability in the near-infrared

In addition to the multi-colour observations (see above) the first observing l period of the survey also included five

A Ks-band observations for each of the 348 tiles. Although the main variability l campaign for the two independent survey areas, i.e. the Bulge and Disc sections, were scheduled for the third and fourth l year of VVV, the search for long-term variables required that at least a few epochs for each tile were obtained in each year/ l

Figure 3. The Bulge region of the VVV survey (196 l tiles, red boxes). The blue dots mark the position of l l known RR Lyrae stars (> 13 000) based on the OGLE ( database.

26 The Messenger 155 – March 2014 510 35 510

30 60 408 408

25 ys ys Da 306 306 Da

20 pochs 40 pochs ulian Julian J

in in . of e . of e

y 15 204 204 y No No Dela Dela 10 20 102 102 5

0 0 0 0 200 250300 350 0 20 40 60 80 10 120140 Tile (Bulge) Tile (Disc)

Figure 4. Distribution of delays between two consec- ing with Baade’s Window, which due to used to create the final 3D model of the utive epochs of the individual Ks-band observations its low Galactic reddening had been inner Milky Way, we have to ensure that for the Bulge (left panel) and the Disc region (right panel) of the VVV survey. The timeline covered by the studied extensively by the Optical Gravi- the individual light curves are indeed Disk observations is February 2010 to October 2013. tational Lensing Experiment (OGLE) pro- of comparable quality, i.e., that the light ject in the optical, and for which a large curves for the variable stars are popu- number of RR Lyrae stars, the survey’s lated in a similar fashion. As for any semester. At the time of this article, the primary targets, are known (see Figure 3). other observing campaign, this not only Disc area had been observed in the depends on the original schedule (e.g., Ks-band 36 times, whereas the Bulge The backbone of the light curve analysis Minniti et al., 2010; Saito et al., 2012b), tiles, subject of the observing campaign is the frequency with which the obser but is also affected by the observing con- in P91, had up to 75 epochs available. vations were executed. This becomes ditions, e.g., presence of clouds, high even more important, when we have to humidity, etc. As a result, the VVV obser- In addition, the observations during Sci- combine data which were not simultane- vations vary significantly in their cadence, ence Verification (e.g., Saito et al., 2010) ously observed, as is the case for the as shown in Figure 4. included additional data for tiles b293 individual VVV tiles. In order to be able to and b294 (a total of 71 epochs), coincid- compare the derived stellar distances Although the observations are still ongoing and will require an additional Figure 5. Examples of 2–3 years, the analysis of the light curves Ks-band light curves of has already begun; they show the excel- long term variables. The lent quality of the data. As shown in Fig- light curves are based on the aperture photom- ures 5 and 6, even at this early stage etry provided by the of the survey, and based on a limited CASU VIRCAM pipeline, number of observing epochs, long-term and use the individual pawprints (Emerson et variables with periods of several hun- al., 2004). dreds of days (Figure 5) can be detected as unambiguously as well as the objects with a period of only a few days or even hours (Figure 6). Only surveys with an anticipated duration period of many years are suitable to study long-term variables such as the ones shown in Figure 5.

Although the VVV survey is targeted at various types of variable stars, of spe- cial importance are those classes of variable stars that show a well-defined luminosity–period correlation, because they will allow us to derive distances and

eventually build a three-dimensional model of the observed Milky Way region. ')#l In addition the frequent observations, in

The Messenger 155 – March 2014 27 Astronomical Science Hempel M. et al., VISTA Variables in the Vía Láctea (VVV)

Figure 6. Examples of six Ks-band light curves of variable stars with periods ranging from a few hours up to several weeks are displayed. As in Figure 6 the aperture photometry of the CASU source catalogues was used.

@F *R L

/G@RD

combination with a limiting Ks magnitude of ~18.0 for a single epoch, allows us to search for and follow up more exotic objects, like novae (Beamin et al., 2013; Saito et al. 2013).

Proper motions

The above-mentioned delays in the observing schedule, which are anything but helpful for the general progress of the survey, can be considered as a blessing in disguise for the proper motion studies, another essential part of the survey (Minniti et al., 2010). Proper motions are the most widely used method to search for close Solar neighbours and thus to complete the census of stars within a few tens of parsecs from the Sun. In particu- lar, in the direction towards the Milky Way Bulge, severe crowding and extinction hamper the detection of faint, albeit nearby stars. (e.g., Lépine et al., 2008);

Figure 7. Selected sample of HPM objects, detected by Gromadzki et al. (2013, see also their Figure 1). The left column shows the 2MASS images, whereas the middle and right columns show the first and last VVV image taken of the object, used to derive the proper motions. Shown from top to bottom are: an HPM M dwarf; an M dwarf + white dwarf common proper motion binary; and a close common proper motion binary.

28 The Messenger 155 – March 2014 these drawbacks can however be DENIS (e.g., Epchtein et al., 1994) allow Technologies PFB-06, the FONDAP Center for overcome by the VVV observations. The us to carry out these studies in an area Astrophysics 15010003 and the MIDEPLAN Milky Way Millennium Nucleus P07-021-F. We would like to 0.339-arcsecond pixels, the much- inaccessible for optical missions, such thank the staff of CASU and WFAU, who provide us reduced dust extinction (in comparison as Gaia. with the pipeline processing, data calibration and with optical observations) and the long archive. Some VVV tiles were made using the Aladin time-line of (eventually) up to seven sky atlas, SExtractor software and products from TERAPIX pipeline. This publication makes use of years (i.e., the full duration of the VVV The future data products from the Two Micron All Sky Survey, a survey) make this survey an ideal tool joint project of the University of Massachusetts and with which to find those missing nearby In recent years, 2MASS has become a IPAC/CALTECH, funded by NASA and the NSF. stars. Already at this early stage of the veritable treasure trove for Milky Way stud- survey, Gromadzki et al. (2013) have ies, and the VVV survey will undoubtedly References found several hundreds of those elusive be its worthy successor for the innermost objects (Figure 7) and, even more impor- regions of the Milky Way. Extending the Beamin, J. C. et al. 2013, ATel, 5215 tantly, while searching only about 31% source lists by applying more sophisti- Catelan, M. et al. 2011, Carnegie Observatories Astrophysics Series, Vol. 5 of the survey area, which corresponds to cated detection algorithms, such as point Dalton, G. B. et al. 2006, Proc. SPIE, 6269 ~1% of the sky. spread function photometry, will allow Emerson, J. & Sutherland, W. 2010, The Messenger, fainter sources to be detected and hence 139, 2 The VVV catalogue of high proper allow distant objects at the far side of Epchtein, N. 1994, A&SS, 217, 3 Gonzalez, O. A. et al. 2011, A&A, 543, L14 motion (HPM) objects includes M dwarfs the Galaxy to be probed. The distances Gonzales, O. A. et al. 2013, The Messenger, 152, 23 towards the Galactic Bulge, common derived for the variable stars within the Gromadzki, M. et al. 2013, Mem. S. A. It., 75, 282 proper motion binaries (see also Ivanov VVV survey area, together with the Ivanov, V. D. et al. 2013, A&A, accepted, et al., 2013), close common proper proper motions, will allow us to build a arXiv:1309.4301v1 Lépine, S. 2008, AJ, 135, 1247 motion M dwarfs + white dwarf binaries model of the mostly unexplored inner Minniti, D. et al. 2010, New Astronomy, 15, 433 and brown dwarfs towards the Galactic regions of the Galaxy. By approximately Minniti, D. et al. 2012, ATel, 4041 Bulge. At this point of the survey proper 2016, the observations for the VVV Nishiyama, S. et al. 2005, ApJ, 621, L105 motions of ~1 arcseconds/yr are studied. survey will be complete and provide the Saito, R. K. et al. 2010, The Messenger, 141, 24 Saito, R. K. et al. 2012a, A&A, 544, 147 Once the late phase of the survey is data required to build a detailed model of Saito, R. K. et al. 2012b, A&A, 537, 107 added, and also the earlier observations the Milky Way Bulge and the southern Saito, R. K. et al. 2013, A&A, 554, 123 with 2MASS (Skrutskie et al., 2006) are Galactic Disc. From there we can take a Schlegel, D. et al. 1998, ApJ, 500, 525 included for the brighter (Ks < 14 mag) large step closer to understanding the Skrutski, M. F. et al. 2006, AJ, 131, 1163 Soto, M. et al. 2013, A&A, 552, 101 objects, an accuracy in the proper motion structure and dynamics of the Galaxy. measurement that almost rivals the one of the Gaia mission (i.e. proper motion Links error: ~10 microarcseconds/yr) will be Acknowledgements 1 VVV stellar density map of the Galactic Bulge: achievable. Further, near-infrared obser- The VVV Survey is supported by ESO, the http://www.eso.org/public/news/eso1242 vations as provided by VVV, 2MASS and BASAL Center for Astrophysics and Associated B. Fugate (FASORtronics)/ESO Fugate B.

Stunning view of the plane of the Milky Way above the Paranal Observatory. At the lower right, the dome of the Residencia reflects the light from the sky and the Large Magellanic Cloud is discernible above.

The Messenger 155 – March 2014 29